Re: bnfisintnorm() bug

[Igor Schein <igor@txc.com>]

On Mon, Jul 01, 2002 at 11:42:46PM +0200, Karim BELABAS wrote:
> On Mon, 1 Jul 2002, Igor Schein wrote:
> > the following enters an infinite loop:
> >
> > bnfisintnorm(bnfinit(x^3+2),5);
> >
> > It was broken some time between 2.2.2 and 2.2.3 releases
> 
> A typo in idealmul(K, prime ideal, principal ideal). Fixed.

On a related note:

   realprecision = 28 significant digits
? tnf=thueinit(x^7-401);
time = 14,200 ms.
?  thue(tnf,88)
  ***   division by zero in gdiv, gdivgs or ginv

Undetected loss of precision.  If I compute tnf at \p140 ( see 
message 1522@pari-dev ), then it works fine.

For comparison:

? bnfisintnorm(bnfinit(x^7-401),88)
  ***   Warning: insufficient precision for fundamental units, not given.
  ***   missing units in bnfnewprec.
  
Insufficient precision properly detected.

Thanks

Igor